Predicting Sleep Using Consumer Wearable Sensing Devices
|
|
- Ralph Blankenship
- 5 years ago
- Views:
Transcription
1 Predicting Sleep Using Consumer Wearable Sensing Devices Miguel A. Garcia Department of Computer Science Stanford University Palo Alto, California 1 Introduction In contrast to the explosion of consumer wearable health device market, which includes millions of units sold worldwide, the efforts undertaken to validate the accuracy and reliability of these monitoring apps and devices have been minimal [1]. To reach this understanding there are several barriers. In the domain of sleep tracking, sleep researchers have yet to define standard metrics for validation, and wearable companies contribute to a general lack of availability of technical information by keeping information about sensors raw data, accuracy, and algorithms hidden as trade secrets [2]. It is increasingly becoming more common for patients to present data acquired from consumer wearables to clinicians (who due to obstacles like those presented before, struggle to interpret this data). Finding reliable devices is a challenge since the acquired data can be at best unregulated and at worst dubious. A widely accepted way to track sleep duration is found in clinical actigraphy, which is a non-invasive method of measuring of motor activity, usually by a small device worn on the wrist. Marino, et al. (2013) validated actigraphy for detecting sleep and wakefulness by using it to yield accurate results resembling those of polysomnography (PSG), which provides the gold standard for this task. However clinical actigraphy devices are expensive ( $1000 per unit). To this end, we study the possibility of employing accelerometer-based consumer wearable devices, which are as widespread as they are affordable (as low as $15). 2 Dataset and Features The dataset for this project was obtained from the Emergent Innovative Global Health Technologies (EIGHT) lab at Stanford. For about 20 different test subjects in the span of 6 months, the Basis Peak watch was used to record 7 different body measurements by the second, including time, heart rate, and galvanic skin response. Since the feature set was already small, when PCA was ran it was not used to project the feature space into a smaller subspace, but rather to analyze what features were most important to categorizing between sleep stages. There were 2 separate datasets used as inputs to the models, both sets of recordings from the same test subject. The first input (dataset A) trained on observations made for the entire month of January 2016, and then tested on observations from the first 2 weeks of February. The second input (dataset B) trained on observations made from January 2016 until March 31, 2016, and then tested on the month of April. Each training example was either unlabeled or labeled with the sleep stage that the test subject was in if they were asleep (deep, rem, or light sleep). This label was also calculated by the watch and can be assumed to be true. 2.1 Pre-processing We re-labeled unlabeled training examples as "awake", scaled individual samples to have unit norm, and filtered out training examples with missing values in any of the features. An extra output vector
2 was made for the binarized task of predicting wakefulness vs general sleep, as all training examples during times of wakefulness were assigned to the value "0", and those during times of sleep in any stage were assigned the value "1". Dataset (A) contained total training examples before being reduced by preprocessing to a final total of 7972 training examples. The test set size for A was after preprocessing. Dataset (B) contained total training examples before being reduced by preprocessing to a final total of training examples. The test set size for B was after preprocessing. After experiencing issues in classification, it was discovered that an imbalance in the dataset in the form of over-representation of wakefulness training examples and under-representation for rem/deep sleep training examples was resulting in flawed results. Thus, an extra step was added to preprocessing to balance the dataset by ensuring there were an equal number of examples for each label. This balancing was not done for the test data. An initial attempt was made at conducting PCA to see whether the feature space could be reduced further, but the explained variance for any feature was not insignificant enough to consider removing it. The following PCA plot was made for our single test subject on dataset A. It is important to note how sleep stages at this point are still inseparable. Figure 1: 2D projection of dataset A (training examples) using PCA. The first principal component is heart rate with an explained variance of 0.44, and the second principal component is galvanic skin response with an explained variance of Models The models are split into 2 different categories for the 2 different tasks in this project: finding a binarized classifier to distinguish between wakefulness and sleep, and finding a multi-class classifier to classify examples based on the different stages of sleep. 3.1 Binary classifiers The labels for this task were "1" for asleep and "0" for awake Logistic Regression Logistic regression works by maximizing the log-likelihood of the training examples. We learn the parameters θ i of the binary model by maximizing l(θ) = m y (i) log(h(x (i) )) + (1 y (i) ) log(1 h(x (i) )) (1) 2
3 3.1.2 Support Vector Machines Support vector machines (SVMs) try to separate data points with a vector while maximizing the margin between training examples of different classes. Given training vectors x i R p, i = 1... n and a vector y {1, 1} n, SVM s can solve the binary classification problem max α m W (α) = α i 1 2 s.t. 0 α i C, i = 1,..., m, m α iy (i) = 0 m y (i) y (j) α i α j K(x (i), x (j) ) (2) i,j=1 The following kernels were used for K: linear : x, x, polynomial : (γ x, x + r) d 3.2 Multi-class classifiers The labels for this task were "awake", "light", "rem", "deep" K-Nearest Neighbors Assign class j to x (i) that maximizes the following: P (y (i) = j x (i) ) = 1 1{y (i) = j} (3) k Softmax Regression Softmax regression is an extension of logistic regression to the multi-class problem. Instead of calculating probabilities for 2 classes, we calculate them for multiple classes. The class predicted is the class with the probability that maximizes the likelihood. We learn the parameters θ i of this model by maximizing ( ) 1{y m k (i) =l} e θt l x(i) l(θ) = log k (4) j=1 eθt j x(i) Support Vector Machines l=1 SVMs can also be used to build a multi-class predictor for k classes by using the "one vs. rest" approach (OVR). In OVR, the algorithm builds an SVM for each class, and the class assigned to each data point is that which maximizes its distance from its respective margin. 4 Experiments The first series of tables below demonstrate the first experiment ran without balancing the dataset. Since this experiment was done without balancing, the training set is larger. While the best binary classifier performs reasonably well, the multi-class classification has artificially high accuracy. Because of the unbalanced dataset, the less frequent stages of sleep "light" and "rem" sleep are almost entirely misclassified. A better measure of performance would be the F1-score, so that we can evaluate the predictor on precision and recall over all classes. The second experiment balanced out the datasets. For the binary case, this meant including as many examples as the sleep state. For the multi-class state, this meant including as many examples as the "deep" sleep state. After looking at the training data I determined these were the least represented categories. There were improvements in the binary case, but in the multi-class task error approached 60% for nearly all the models. Here are the results below for the binary case (with k-nearest neighbors performed this time around): knn train err: % knn dev err: % knn test err: % i N 3
4 Classifier Train Dev Test Logistic Reg. (L2 penalty) 6.99% 7.32% 8.86% SVM (linear) 6.83% 7.19% 9.15% SVM (d = 2) 9.17% 9.68% 11.5% SVM (d = 3) 6.61% 7.09% 8.09% Table 1: Binary classifier error (hold-out cross validation) Classifier Train Dev Test K-Nearest Neighbors 9.03% 13.43% 20.66% Softmax Reg % 17.25% 15.03% Linear SVM 17.31% 16.88% 15.42% SVM (d=2) 19.18% 19.03% 17.28% SVM (d=3) 17.13% 16.91% 14.94% Table 2: Multi-class classifier error (hold-out cross validation) awake asleep awake asleep Table 3: Actual vs predicted for binary SVM(d=3) awake rem light deep awake rem light deep Table 4: Actual vs predicted for multi-class SVM(d=3) Test set f1 score: LR train err: % LR dev err: % LR test err: % Test set f1 score: SVM (linear, 0) train err: % SVM (linear, 0) dev err: % SVM (linear, 0) test err: % Test set f1 score: SVM (poly, 2) train err: % SVM (poly, 2) dev err: % SVM (poly, 2) test err: % Test set f1 score: SVM (poly, 3) train err: % SVM (poly, 3) dev err: % SVM (poly, 3) test err: % Test set f1 score:
5 5 Discussion Although the results are acceptable for the binary case, there is much room for improvement in the multi-class state. Since the models do not generalize well, using more data might improve them. However, much of the models are also hampered by the limited feature set. Perhaps adding more features by collecting more information about the user with the Basis Peak watch might help, though data collected by another device, such as a smartphone, could be combined someway with the data we have collected already as well. 6 Future Work With more time, I wold have liked to improve upon these models. Much of my time was spent understanding the models and learning how to evaluate the results of the experiments. Once the models are improved, I would like to use a mixture of Guassians model to combine data from multiple people, and see whether these distributions fit the Gaussian model. Acknowledgments I would like to thank the EIGHT lab for their help in providing me with their datasets and Dr. Katarzyna Wac for her guidance. References [1] de Zambotti M, Godino JG, Baker FC, Cheung J, Patrick K, Colrain IM. The boom in wearable technology: Cause for alarm or just what is needed to better understand sleep? Sleep. 2016;In Press: doi: /sleep [2] Marino M, Li Y, Rueschman MN, Winkelman JW, Ellenbogen JM, Solet JM, Dulin H, Berkman LF, Buxton OM. Measuring Sleep: Accuracy, Sensitivity, and Specificity of Wrist Actigraphy Compared to Polysomnography. Sleep. 2013;36(11): doi: /sleep [4] Hasselmo, M.E., Schnell, E. & Barkai, E. (1995) Dynamics of learning and recall at excitatory recurrent synapses and cholinergic modulation in rat hippocampal region CA3. Journal of Neuroscience 15(7):
Wearable Technology and Apps Alanna Cornish, BSc., RPSGT, RST
Wearable Technology and Apps Alanna Cornish, BSc., RPSGT, RST There are many wearable devices and apps available on the market to track sleep duration and quality. With more and more patients accessing
More informationActigraphy-based Clinical Study Endpoints: A Regulatory Perspective
Actigraphy-based Clinical Study Endpoints: A Regulatory Perspective Ebony Dashiell-Aje, PhD Clinical Outcome Assessments Staff Office of New Drugs Center for Drug Evaluation and Research U.S. Food and
More informationAssigning B cell Maturity in Pediatric Leukemia Gabi Fragiadakis 1, Jamie Irvine 2 1 Microbiology and Immunology, 2 Computer Science
Assigning B cell Maturity in Pediatric Leukemia Gabi Fragiadakis 1, Jamie Irvine 2 1 Microbiology and Immunology, 2 Computer Science Abstract One method for analyzing pediatric B cell leukemia is to categorize
More informationData mining for Obstructive Sleep Apnea Detection. 18 October 2017 Konstantinos Nikolaidis
Data mining for Obstructive Sleep Apnea Detection 18 October 2017 Konstantinos Nikolaidis Introduction: What is Obstructive Sleep Apnea? Obstructive Sleep Apnea (OSA) is a relatively common sleep disorder
More informationUNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Final, Fall 2014
UNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Final, Fall 2014 Exam policy: This exam allows two one-page, two-sided cheat sheets (i.e. 4 sides); No other materials. Time: 2 hours. Be sure to write
More informationRecognition of Sleep Dependent Memory Consolidation with Multi-modal Sensor Data
Recognition of Sleep Dependent Memory Consolidation with Multi-modal Sensor Data The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation
More informationEECS 433 Statistical Pattern Recognition
EECS 433 Statistical Pattern Recognition Ying Wu Electrical Engineering and Computer Science Northwestern University Evanston, IL 60208 http://www.eecs.northwestern.edu/~yingwu 1 / 19 Outline What is Pattern
More informationMachine Learning for Predicting Delayed Onset Trauma Following Ischemic Stroke
Machine Learning for Predicting Delayed Onset Trauma Following Ischemic Stroke Anthony Ma 1, Gus Liu 1 Department of Computer Science, Stanford University, Stanford, CA 94305 Stroke is currently the third
More informationBREAST CANCER EPIDEMIOLOGY MODEL:
BREAST CANCER EPIDEMIOLOGY MODEL: Calibrating Simulations via Optimization Michael C. Ferris, Geng Deng, Dennis G. Fryback, Vipat Kuruchittham University of Wisconsin 1 University of Wisconsin Breast Cancer
More informationA Vision-based Affective Computing System. Jieyu Zhao Ningbo University, China
A Vision-based Affective Computing System Jieyu Zhao Ningbo University, China Outline Affective Computing A Dynamic 3D Morphable Model Facial Expression Recognition Probabilistic Graphical Models Some
More informationIntroduction to Machine Learning. Katherine Heller Deep Learning Summer School 2018
Introduction to Machine Learning Katherine Heller Deep Learning Summer School 2018 Outline Kinds of machine learning Linear regression Regularization Bayesian methods Logistic Regression Why we do this
More informationTHE data used in this project is provided. SEIZURE forecasting systems hold promise. Seizure Prediction from Intracranial EEG Recordings
1 Seizure Prediction from Intracranial EEG Recordings Alex Fu, Spencer Gibbs, and Yuqi Liu 1 INTRODUCTION SEIZURE forecasting systems hold promise for improving the quality of life for patients with epilepsy.
More informationError Detection based on neural signals
Error Detection based on neural signals Nir Even- Chen and Igor Berman, Electrical Engineering, Stanford Introduction Brain computer interface (BCI) is a direct communication pathway between the brain
More informationPredicting Diabetes and Heart Disease Using Features Resulting from KMeans and GMM Clustering
Predicting Diabetes and Heart Disease Using Features Resulting from KMeans and GMM Clustering Kunal Sharma CS 4641 Machine Learning Abstract Clustering is a technique that is commonly used in unsupervised
More informationApplying One-vs-One and One-vs-All Classifiers in k-nearest Neighbour Method and Support Vector Machines to an Otoneurological Multi-Class Problem
Oral Presentation at MIE 2011 30th August 2011 Oslo Applying One-vs-One and One-vs-All Classifiers in k-nearest Neighbour Method and Support Vector Machines to an Otoneurological Multi-Class Problem Kirsi
More informationClass discovery in Gene Expression Data: Characterizing Splits by Support Vector Machines
Class discovery in Gene Expression Data: Characterizing Splits by Support Vector Machines Florian Markowetz and Anja von Heydebreck Max-Planck-Institute for Molecular Genetics Computational Molecular Biology
More informationMachine Learning to Inform Breast Cancer Post-Recovery Surveillance
Machine Learning to Inform Breast Cancer Post-Recovery Surveillance Final Project Report CS 229 Autumn 2017 Category: Life Sciences Maxwell Allman (mallman) Lin Fan (linfan) Jamie Kang (kangjh) 1 Introduction
More information10CS664: PATTERN RECOGNITION QUESTION BANK
10CS664: PATTERN RECOGNITION QUESTION BANK Assignments would be handed out in class as well as posted on the class blog for the course. Please solve the problems in the exercises of the prescribed text
More informationCollaborating to Develop Digital Biomarkers with Passive Data Collection
Collaborating to Develop Digital Biomarkers with Passive Data Collection Iain Simpson IXICO June 2018 1 Setting of Data Collection Market evolution: biosensors and digital biomarkers Clinic Home Digital
More informationChallenges in Developing Learning Algorithms to Personalize mhealth Treatments
Challenges in Developing Learning Algorithms to Personalize mhealth Treatments JOOLHEALTH Bar-Fit Susan A Murphy 01.16.18 HeartSteps SARA Sense 2 Stop Continually Learning Mobile Health Intervention 1)
More informationVital Responder: Real-time Health Monitoring of First- Responders
Vital Responder: Real-time Health Monitoring of First- Responders Ye Can 1,2 Advisors: Miguel Tavares Coimbra 2, Vijayakumar Bhagavatula 1 1 Department of Electrical & Computer Engineering, Carnegie Mellon
More informationClassification of EEG signals in an Object Recognition task
Classification of EEG signals in an Object Recognition task Iacob D. Rus, Paul Marc, Mihaela Dinsoreanu, Rodica Potolea Technical University of Cluj-Napoca Cluj-Napoca, Romania 1 rus_iacob23@yahoo.com,
More informationData Mining in Bioinformatics Day 4: Text Mining
Data Mining in Bioinformatics Day 4: Text Mining Karsten Borgwardt February 25 to March 10 Bioinformatics Group MPIs Tübingen Karsten Borgwardt: Data Mining in Bioinformatics, Page 1 What is text mining?
More informationMachine learning for detection of epileptic seizures
Machine learning for detection of epileptic seizures Evaluation of features and machine learning methods for acceleration based classification of generalized tonic-clonic seizures Master s thesis in Systems,
More informationReview: Logistic regression, Gaussian naïve Bayes, linear regression, and their connections
Review: Logistic regression, Gaussian naïve Bayes, linear regression, and their connections New: Bias-variance decomposition, biasvariance tradeoff, overfitting, regularization, and feature selection Yi
More informationMethods for Predicting Type 2 Diabetes
Methods for Predicting Type 2 Diabetes CS229 Final Project December 2015 Duyun Chen 1, Yaxuan Yang 2, and Junrui Zhang 3 Abstract Diabetes Mellitus type 2 (T2DM) is the most common form of diabetes [WHO
More informationNMF-Density: NMF-Based Breast Density Classifier
NMF-Density: NMF-Based Breast Density Classifier Lahouari Ghouti and Abdullah H. Owaidh King Fahd University of Petroleum and Minerals - Department of Information and Computer Science. KFUPM Box 1128.
More informationIdentifying Thyroid Carcinoma Subtypes and Outcomes through Gene Expression Data Kun-Hsing Yu, Wei Wang, Chung-Yu Wang
Identifying Thyroid Carcinoma Subtypes and Outcomes through Gene Expression Data Kun-Hsing Yu, Wei Wang, Chung-Yu Wang Abstract: Unlike most cancers, thyroid cancer has an everincreasing incidence rate
More informationDiscovering Meaningful Cut-points to Predict High HbA1c Variation
Proceedings of the 7th INFORMS Workshop on Data Mining and Health Informatics (DM-HI 202) H. Yang, D. Zeng, O. E. Kundakcioglu, eds. Discovering Meaningful Cut-points to Predict High HbAc Variation Si-Chi
More informationConfluence: Conformity Influence in Large Social Networks
Confluence: Conformity Influence in Large Social Networks Jie Tang *, Sen Wu *, and Jimeng Sun + * Tsinghua University + IBM TJ Watson Research Center 1 Conformity Conformity is the act of matching attitudes,
More informationPredicting Seizures in Intracranial EEG Recordings
Sining Ma, Jiawei Zhu sma87@stanford.edu, jiaweiz@stanford.edu Abstract If seizure forecasting systems could reliably identify periods of increased probability of seizure occurrence, patients who suffer
More informationUNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Midterm, 2016
UNIVERSITY of PENNSYLVANIA CIS 520: Machine Learning Midterm, 2016 Exam policy: This exam allows one one-page, two-sided cheat sheet; No other materials. Time: 80 minutes. Be sure to write your name and
More informationPredicting About-to-Eat Moments for Just-in-Time Eating Intervention
Predicting About-to-Eat Moments for Just-in-Time Eating Intervention CORNELL UNIVERSITY AND VIBE GROUP AT MICROSOFT RESEARCH Motivation Obesity is a leading cause of preventable death second only to smoking,
More informationSleep, Optimized. Accurate multistage sleep tracking for a new level of personal performance
Sleep, Optimized Accurate multistage sleep tracking for a new level of personal performance Experience Optimized Sleep In a world that increasingly demands more from us, accomplishing your daytime goals
More informationTITLE: A Data-Driven Approach to Patient Risk Stratification for Acute Respiratory Distress Syndrome (ARDS)
TITLE: A Data-Driven Approach to Patient Risk Stratification for Acute Respiratory Distress Syndrome (ARDS) AUTHORS: Tejas Prahlad INTRODUCTION Acute Respiratory Distress Syndrome (ARDS) is a condition
More informationSparsifying machine learning models identify stable subsets of predictive features for behavioral detection of autism
Levy et al. Molecular Autism (2017) 8:65 DOI 10.1186/s13229-017-0180-6 RESEARCH Sparsifying machine learning models identify stable subsets of predictive features for behavioral detection of autism Sebastien
More informationLarge-Scale Statistical Modelling via Machine Learning Classifiers
J. Stat. Appl. Pro. 2, No. 3, 203-222 (2013) 203 Journal of Statistics Applications & Probability An International Journal http://dx.doi.org/10.12785/jsap/020303 Large-Scale Statistical Modelling via Machine
More informationClassification of Epileptic Seizure Predictors in EEG
Classification of Epileptic Seizure Predictors in EEG Problem: Epileptic seizures are still not fully understood in medicine. This is because there is a wide range of potential causes of epilepsy which
More informationNeuroinformatics. Ilmari Kurki, Urs Köster, Jukka Perkiö, (Shohei Shimizu) Interdisciplinary and interdepartmental
Neuroinformatics Aapo Hyvärinen, still Academy Research Fellow for a while Post-docs: Patrik Hoyer and Jarmo Hurri + possibly international post-docs PhD students Ilmari Kurki, Urs Köster, Jukka Perkiö,
More informationNature Neuroscience: doi: /nn Supplementary Figure 1. Behavioral training.
Supplementary Figure 1 Behavioral training. a, Mazes used for behavioral training. Asterisks indicate reward location. Only some example mazes are shown (for example, right choice and not left choice maze
More informationPredicting Kidney Cancer Survival from Genomic Data
Predicting Kidney Cancer Survival from Genomic Data Christopher Sauer, Rishi Bedi, Duc Nguyen, Benedikt Bünz Abstract Cancers are on par with heart disease as the leading cause for mortality in the United
More informationComputer Age Statistical Inference. Algorithms, Evidence, and Data Science. BRADLEY EFRON Stanford University, California
Computer Age Statistical Inference Algorithms, Evidence, and Data Science BRADLEY EFRON Stanford University, California TREVOR HASTIE Stanford University, California ggf CAMBRIDGE UNIVERSITY PRESS Preface
More informationVariable Features Selection for Classification of Medical Data using SVM
Variable Features Selection for Classification of Medical Data using SVM Monika Lamba USICT, GGSIPU, Delhi, India ABSTRACT: The parameters selection in support vector machines (SVM), with regards to accuracy
More informationImproved Intelligent Classification Technique Based On Support Vector Machines
Improved Intelligent Classification Technique Based On Support Vector Machines V.Vani Asst.Professor,Department of Computer Science,JJ College of Arts and Science,Pudukkottai. Abstract:An abnormal growth
More informationAn Improved Algorithm To Predict Recurrence Of Breast Cancer
An Improved Algorithm To Predict Recurrence Of Breast Cancer Umang Agrawal 1, Ass. Prof. Ishan K Rajani 2 1 M.E Computer Engineer, Silver Oak College of Engineering & Technology, Gujarat, India. 2 Assistant
More informationApps and fitness trackers that measure sleep: Are they useful?
REVIEW MEGHNA P. MANSUKHANI, MD Center for Sleep Medicine, Mayo Clinic, Rochester, MN BHANU PRAKASH KOLLA, MD, MRCPsych Center for Sleep Medicine and Department of Psychiatry and Psychology, Mayo Clinic,
More informationSupporting Information Identification of Amino Acids with Sensitive Nanoporous MoS 2 : Towards Machine Learning-Based Prediction
Supporting Information Identification of Amino Acids with Sensitive Nanoporous MoS 2 : Towards Machine Learning-Based Prediction Amir Barati Farimani, Mohammad Heiranian, Narayana R. Aluru 1 Department
More informationPart [2.1]: Evaluation of Markers for Treatment Selection Linking Clinical and Statistical Goals
Part [2.1]: Evaluation of Markers for Treatment Selection Linking Clinical and Statistical Goals Patrick J. Heagerty Department of Biostatistics University of Washington 174 Biomarkers Session Outline
More informationCLASSIFICATION OF BREAST CANCER INTO BENIGN AND MALIGNANT USING SUPPORT VECTOR MACHINES
CLASSIFICATION OF BREAST CANCER INTO BENIGN AND MALIGNANT USING SUPPORT VECTOR MACHINES K.S.NS. Gopala Krishna 1, B.L.S. Suraj 2, M. Trupthi 3 1,2 Student, 3 Assistant Professor, Department of Information
More informationSleep Stage Estimation By Evolutionary Computation Using Heartbeat Data and Body-Movement
Sleep Stage Estimation By Evolutionary Computation Using Heartbeat Data and Body-Movement 1 Hiroyasu MAtsushima, 2 Kazuyuki Hirose, 3 Kiyohiko Hattori, 4 Hiroyuki Sato, 5 Keiki Takadama 1-5, First Author
More informationStatistics 202: Data Mining. c Jonathan Taylor. Final review Based in part on slides from textbook, slides of Susan Holmes.
Final review Based in part on slides from textbook, slides of Susan Holmes December 5, 2012 1 / 1 Final review Overview Before Midterm General goals of data mining. Datatypes. Preprocessing & dimension
More informationIdentifying Parkinson s Patients: A Functional Gradient Boosting Approach
Identifying Parkinson s Patients: A Functional Gradient Boosting Approach Devendra Singh Dhami 1, Ameet Soni 2, David Page 3, and Sriraam Natarajan 1 1 Indiana University Bloomington 2 Swarthmore College
More informationComputational Cognitive Neuroscience
Computational Cognitive Neuroscience Computational Cognitive Neuroscience Computational Cognitive Neuroscience *Computer vision, *Pattern recognition, *Classification, *Picking the relevant information
More informationGene Selection for Tumor Classification Using Microarray Gene Expression Data
Gene Selection for Tumor Classification Using Microarray Gene Expression Data K. Yendrapalli, R. Basnet, S. Mukkamala, A. H. Sung Department of Computer Science New Mexico Institute of Mining and Technology
More informationSleepMonitor: Monitoring Respiratory Rate and Body Position During Sleep Using Smartwatch
SleepMonitor: Monitoring Respiratory Rate and Body Position During Sleep Using Smartwatch XIAO SUN, LI QIU, YIBO WU, YEMING TANG, and GUOHONG CAO, The Pennsylvania State University, University Park. Respiratory
More informationThe Long Tail of Recommender Systems and How to Leverage It
The Long Tail of Recommender Systems and How to Leverage It Yoon-Joo Park Stern School of Business, New York University ypark@stern.nyu.edu Alexander Tuzhilin Stern School of Business, New York University
More informationBayesOpt: Extensions and applications
BayesOpt: Extensions and applications Javier González Masterclass, 7-February, 2107 @Lancaster University Agenda of the day 9:00-11:00, Introduction to Bayesian Optimization: What is BayesOpt and why it
More informationEvaluating Classifiers for Disease Gene Discovery
Evaluating Classifiers for Disease Gene Discovery Kino Coursey Lon Turnbull khc0021@unt.edu lt0013@unt.edu Abstract Identification of genes involved in human hereditary disease is an important bioinfomatics
More informationAutomatic Classification of Perceived Gender from Facial Images
Automatic Classification of Perceived Gender from Facial Images Joseph Lemley, Sami Abdul-Wahid, Dipayan Banik Advisor: Dr. Razvan Andonie SOURCE 2016 Outline 1 Introduction 2 Faces - Background 3 Faces
More informationPredicting Sleeping Behaviors in Long-Term Studies with Wrist-Worn Sensor Data
Predicting Sleeping Behaviors in Long-Term Studies with Wrist-Worn Sensor Data Marko Borazio and Kristof Van Laerhoven TU-Darmstadt, Germany http://www.ess.tu-darmstadt.de Abstract. This paper conducts
More informationAnalysis of Rheumatoid Arthritis Data using Logistic Regression and Penalized Approach
University of South Florida Scholar Commons Graduate Theses and Dissertations Graduate School November 2015 Analysis of Rheumatoid Arthritis Data using Logistic Regression and Penalized Approach Wei Chen
More informationDesign of Multi-Class Classifier for Prediction of Diabetes using Linear Support Vector Machine
Design of Multi-Class Classifier for Prediction of Diabetes using Linear Support Vector Machine Akshay Joshi Anum Khan Omkar Kulkarni Department of Computer Engineering Department of Computer Engineering
More informationBeating Diabetes: Predicting Early Diabetes Patient Hospital Readmittance to Help Optimize Patient Care
Beating Diabetes: Predicting Early Diabetes Patient Hospital Readmittance to Help Optimize Patient Care Project Category: Life Sciences Introduction According to the American Society of Diabetes, the cost
More informationJ2.6 Imputation of missing data with nonlinear relationships
Sixth Conference on Artificial Intelligence Applications to Environmental Science 88th AMS Annual Meeting, New Orleans, LA 20-24 January 2008 J2.6 Imputation of missing with nonlinear relationships Michael
More informationAlways, Everywhere, Never Alone. Saving lives with ultra accurate technology. For athletes. Training and monitoring
Always, Everywhere, Never Alone Saving lives with ultra accurate technology For athletes. Training and monitoring What is? Lightbringer is an innovative system able to track biometric data through ultra-accurate
More informationCognitive Health Prediction on the Elderly Using Sensor Data in Smart Homes
The Thirty-First International Florida Artificial Intelligence Research Society Conference (FLAIRS-31) Cognitive Health Prediction on the Elderly Using Sensor Data in Smart Homes Ramesh Paudel, 1 Kimberlyn
More informationFUNNEL: Automatic Mining of Spatially Coevolving Epidemics
FUNNEL: Automatic Mining of Spatially Coevolving Epidemics By Yasuo Matsubara, Yasushi Sakurai, Willem G. van Panhuis, and Christos Faloutsos SIGKDD 2014 Presented by Sarunya Pumma This presentation has
More informationGIANT: Geo-Informative Attributes for Location Recognition and Exploration
GIANT: Geo-Informative Attributes for Location Recognition and Exploration Quan Fang, Jitao Sang, Changsheng Xu Institute of Automation, Chinese Academy of Sciences October 23, 2013 Where is this? La Sagrada
More informationThe Sleep of the Ring: Comparison of the ŌURA Sleep Tracker Against Polysomnography
Behavioral Sleep Medicine ISSN: 1540-2002 (Print) 1540-2010 (Online) Journal homepage: http://www.tandfonline.com/loi/hbsm20 The Sleep of the Ring: Comparison of the ŌURA Sleep Tracker Against Polysomnography
More informationPredicting Breast Cancer Survival Using Treatment and Patient Factors
Predicting Breast Cancer Survival Using Treatment and Patient Factors William Chen wchen808@stanford.edu Henry Wang hwang9@stanford.edu 1. Introduction Breast cancer is the leading type of cancer in women
More informationSleepGuard: Capturing Rich Sleep Information Using Smartwatch Sensing Data
98 SleepGuard: Capturing Rich Sleep Information Using Smartwatch Sensing Data LIQIONG CHANG, JIAQI LU, JU WANG, XIAOJIANG CHEN, and DINGYI FANG, ZHANY- ONG TANG, Northwest University, China PETTERI NURMI,
More informationInferring Clinical Correlations from EEG Reports with Deep Neural Learning
Inferring Clinical Correlations from EEG Reports with Deep Neural Learning Methods for Identification, Classification, and Association using EHR Data S23 Travis R. Goodwin (Presenter) & Sanda M. Harabagiu
More informationGender Based Emotion Recognition using Speech Signals: A Review
50 Gender Based Emotion Recognition using Speech Signals: A Review Parvinder Kaur 1, Mandeep Kaur 2 1 Department of Electronics and Communication Engineering, Punjabi University, Patiala, India 2 Department
More informationSleep Stages Solution v0.1
Sleep Stages Solution v0.1 June 2016 Table of contents Key terms... 2 Introduction... 2 Test Protocol... 3 Inputs and Outputs... 4 Validation and Accuracy... 4 Example datasets... 6 Limitations of the
More informationPredicting Breast Cancer Recurrence Using Machine Learning Techniques
Predicting Breast Cancer Recurrence Using Machine Learning Techniques Umesh D R Department of Computer Science & Engineering PESCE, Mandya, Karnataka, India Dr. B Ramachandra Department of Electrical and
More informationWHO WE ARE WHO USES MIO MIO S PRODUCTS INCLUDE WHAT WE DO WHY TRAIN WITH HEART. Mio FUSE. Mio ALPHA 2. Mio LINK
WHO WE ARE Mio Global makes advanced wearable technology for athletes who want to improve their performance and get the most out of each training session. The company was founded in 1999 by CEO Liz Dickinson,
More informationHIWIN Thesis Award 2007
HIWIN Thesis Award 2007 Optimal Design Laboratory & Gerontechnology Research Center Yuan Ze University Physical Activity Physical activity can be regarded as any bodily movement or posture produced by
More informationAn Edge-Device for Accurate Seizure Detection in the IoT
An Edge-Device for Accurate Seizure Detection in the IoT M. A. Sayeed 1, S. P. Mohanty 2, E. Kougianos 3, and H. Zaveri 4 University of North Texas, Denton, TX, USA. 1,2,3 Yale University, New Haven, CT,
More informationAugmented Medical Decisions
Machine Learning Applied to Biomedical Challenges 2016 Rulex, Inc. Intelligible Rules for Reliable Diagnostics Rulex is a predictive analytics platform able to manage and to analyze big amounts of heterogeneous
More informationClassifying Adolescent Excessive Alcohol Drinkers from fmri Data
CLASSIFYING ADOLESCENT EXCESSIVE ALCOHOL DRINKERS FROM FMRI DATA 1 Classifying Adolescent Excessive Alcohol Drinkers from fmri Data Yong-hun Kim (ykim9), Cindy Liu (cliu15), Joseph Noh (jnoh2) Abstract
More informationUtilizing Posterior Probability for Race-composite Age Estimation
Utilizing Posterior Probability for Race-composite Age Estimation Early Applications to MORPH-II Benjamin Yip NSF-REU in Statistical Data Mining and Machine Learning for Computer Vision and Pattern Recognition
More informationWhite Paper Estimating Complex Phenotype Prevalence Using Predictive Models
White Paper 23-12 Estimating Complex Phenotype Prevalence Using Predictive Models Authors: Nicholas A. Furlotte Aaron Kleinman Robin Smith David Hinds Created: September 25 th, 2015 September 25th, 2015
More informationSound Texture Classification Using Statistics from an Auditory Model
Sound Texture Classification Using Statistics from an Auditory Model Gabriele Carotti-Sha Evan Penn Daniel Villamizar Electrical Engineering Email: gcarotti@stanford.edu Mangement Science & Engineering
More informationWinner s Report: KDD CUP Breast Cancer Identification
Winner s Report: KDD CUP Breast Cancer Identification ABSTRACT Claudia Perlich, Prem Melville, Yan Liu, Grzegorz Świrszcz, Richard Lawrence IBM T.J. Watson Research Center Yorktown Heights, NY 10598 {perlich,pmelvil,liuya}@us.ibm.com
More informationEnhanced Detection of Lung Cancer using Hybrid Method of Image Segmentation
Enhanced Detection of Lung Cancer using Hybrid Method of Image Segmentation L Uma Maheshwari Department of ECE, Stanley College of Engineering and Technology for Women, Hyderabad - 500001, India. Udayini
More information'Automated dermatologist' detects skin cancer with expert accuracy - CNN.com
'Automated dermatologist' detects skin cancer with expert accuracy (CNN)Even though the phrase "image recognition technologies" conjures visions of high-tech surveillance, these tools may soon be used
More informationSmartwatches & Biometrics for Personal Health Monitoring. Josh Piemontesi, Braden Siempelkamp, and Gordon Minaker
Smartwatches & Biometrics for Personal Health Monitoring Josh Piemontesi, Braden Siempelkamp, and Gordon Minaker Disclosures Nothing to disclose. No potential conflict of interest. No companies paying
More informationPredicting Breast Cancer Survivability Rates
Predicting Breast Cancer Survivability Rates For data collected from Saudi Arabia Registries Ghofran Othoum 1 and Wadee Al-Halabi 2 1 Computer Science, Effat University, Jeddah, Saudi Arabia 2 Computer
More informationPersonalized Effect of Health Behavior on Blood Pressure: Machine Learning Based Prediction and Recommendation
Personalized Effect of Health Behavior on Blood Pressure: Machine Learning Based Prediction and Recommendation Po-Han Chiang and Sujit Dey Mobile Systems Design Lab, Dept. of Electrical and Computer Engineering,
More informationSeizure Prediction and Detection
Seizure Prediction and Detection Tay Netoff Yun-Sang Park Michael Brown Lan Luo University of Minnesota Minneapolis, MN Feb. 9, 2011 Outline Seizure prediction using spectral power features and SVMclassification
More informationABSTRACT I. INTRODUCTION. Mohd Thousif Ahemad TSKC Faculty Nagarjuna Govt. College(A) Nalgonda, Telangana, India
International Journal of Scientific Research in Computer Science, Engineering and Information Technology 2018 IJSRCSEIT Volume 3 Issue 1 ISSN : 2456-3307 Data Mining Techniques to Predict Cancer Diseases
More informationClassification of Honest and Deceitful Memory in an fmri Paradigm CS 229 Final Project Tyler Boyd Meredith
12/14/12 Classification of Honest and Deceitful Memory in an fmri Paradigm CS 229 Final Project Tyler Boyd Meredith Introduction Background and Motivation In the past decade, it has become popular to use
More informationSmartphone Applications to Support Sleep Self-Management
Smartphone Applications to Support Sleep Self-Management Dr Pierre El Chater 06/12/18 SOMNOFORUM BERLIN 2018 1 Assessing available technologies in the market for sleep management and its effectivness Dr
More informationPredictive performance and discrimination in unbalanced classification
MASTER Predictive performance and discrimination in unbalanced classification van der Zon, S.B. Award date: 2016 Link to publication Disclaimer This document contains a student thesis (bachelor's or master's),
More informationA Classification of Polycystic Ovary Syndrome Based on Follicle Detection of Ultrasound Images
A Classification of Polycystic Ovary Syndrome Based on Follicle Detection of Ultrasound Images Bedy Purnama 1, Untari Novia Wisesti 2, Adiwijaya 3, Fhira Nhita 4, Andini Gayatri 5, Titik Mutiah 6 School
More informationBACKPROPOGATION NEURAL NETWORK FOR PREDICTION OF HEART DISEASE
BACKPROPOGATION NEURAL NETWORK FOR PREDICTION OF HEART DISEASE NABEEL AL-MILLI Financial and Business Administration and Computer Science Department Zarqa University College Al-Balqa' Applied University
More informationNeural Network for Detecting Head Impacts from Kinematic Data. Michael Fanton, Nicholas Gaudio, Alissa Ling CS 229 Project Report
Neural Network for Detecting Head Impacts from Kinematic Data Michael Fanton, Nicholas Gaudio, Alissa Ling CS 229 Project Report 1. Abstract Mild Traumatic Brain Injury (mtbi) is a serious health concern,
More informationDiagnosis of Breast Cancer Using Ensemble of Data Mining Classification Methods
International Journal of Bioinformatics and Biomedical Engineering Vol. 1, No. 3, 2015, pp. 318-322 http://www.aiscience.org/journal/ijbbe ISSN: 2381-7399 (Print); ISSN: 2381-7402 (Online) Diagnosis of
More information